Nuclear Inst.

and Methods in Physics Research, A 889 (2018) 118–121

Contents lists available at ScienceDirect

Nuclear Inst. and Methods in Physics Research, A

journal homepage: www.elsevier.com/locate/nima

Penetration of the consolidant Paraloid® B-72 in Macuxi indigenous

ceramic vessels investigated by neutron tomography
Marco A. Stanojev Pereira, Reynaldo Pugliesi *
Instituto de Pesquisas Energéticas e Nucleares, Centro do Reator de Pesquisas. Av. Lineu Prestes 2242, Cidade Universitária, 05508-000 São Paulo, Brazil

a r t i c l e i n f o a b s t r a c t

Keywords: The neutron tomography technique was applied in studying the penetration of the consolidant Paraloid® B-72 in
Neutron tomography contemporary indigenous ceramic vessels. The study was carried out for two distinct and controlled air humidity
Cultural heritage objects
conditions, 40% and 90%, in which the vessels were exposed, before the consolidant application. The obtained
Ceramic restorations
images have proved that the penetration of Paraloid® B-72 in the ceramic does not depend on the humidity
condition in which it was applied, moreover allowed a macro-visualization of the consolidant penetration in the
ceramic vessel. As the vessels used in the present work were manufactured by an indigenous artisan, Macuxi,
according to the same procedures and raw materials used by the ancient artisans, the results obtained can be
used as a guide to assist experts, both in the study of archeological objects of Macuxi origin, as well as other
objects that had been made by other tribes that lived in the same Amazon region, in Brazil.

1. Introduction penetration as well as the distribution of Paraloid® B-72 within the

Objects manufactured out of clay are present in many cultures over
several epochs of civilization. In Brazil, ceramic objects date back to
2000 BC, especially those produced on Marajó Island [1,2]. Before the
European arrival in America, there were about 5 million natives living
in the Brazilian territory. They were divided in tribes according to the
linguistic affiliation in Tupi–Guarani (Coast), Macro-jê (Central plateau),
Aruak and Karib (Amazon). Particularly, the Macuxi tribe speaks a
language derived from the Karib family, and since the middle of the
XVII century, they live in the northeast of the Roraima State, between
the Branco and Rupununi river heads, on the border of Brazil and
Guiana. Besides the Macuxis, other tribes like Yanomami, Ingarikó, Wai-
wai, Patamona, Wapixana, Waimiri-atroari, Yekuana, Taurepang, Arekuna
and Kamarakoto, lived in the same region, and because of the closeness,
the tribes are social and culturally alike [3]. Their rich cultural heritage
includes, among others, the manufacture of objects, and for the purposes
of the present work, the ceramic ones will be emphasized.
In order to preserve archeological ceramic objects, it is usual the em-
ployment of restoration techniques to prevent deterioration caused by
decomposition or fragmentation processes. In such cases, the restorers
usually make use of specific organic substances known as consolidants,
which are applied to the surfaces of the objects to be restored. These
substances are acrylic polymers that, after application, penetrate by the
pores of the ceramic forming a protective film on its surface. Both the
structure of the object of interest are key factors to the effectiveness
of the employed restoration technique [4–8].
The objective of the present study was to investigate the penetration
and the distribution of the consolidant Paraloid® B-72 in Macuxi
contemporary indigenous ceramic vessels, by means of the neutron
tomography technique. The consolidant was applied in two humidity
conditions in which archeological objects are usually found in the
Brazilian Amazon [9].
2. The ceramic samples and the consolidant
The ceramic samples studied in the present work were contemporary
vessels manufactured by an artisan of the Brazilian tribe Macuxi, making
use of the same manufacturing process and raw-materials used by their
ancestors. The raw-material is clay which, after being collected close to
the indigenous settlement, is dried in the environment and crushed using
a wooden pestle. The crushed clay is slightly moistened and the vessels
are hand molded and dried in an ordinary campfire [9]. The resulting
vessels are not uniform both in shape as well as in thickness. Their
maximal external diameters vary from 85 to 90 mm while the thickness
of the walls, from 5 to 9 mm. Typical vessels are shown in Fig. 1.
The consolidant Paraloid® B-72 is an ethyl methacrylate (70%) and
methyl acrylate (30%) copolymer which, for the purposes of the present
work, usually is diluted in acetone P.A. 10% in mass [4–8].

* Corresponding author.
E-mail address: pugliesi@ipen.br (R. Pugliesi).

https://doi.org/10.1016/j.nima.2018.02.034
Received 5 October 2017; Received in revised form 2 February 2018; Accepted 6 February 2018
Available online 9 February 2018
0168-9002/© 2018 Elsevier B.V. All rights reserved.
M.A. Stanojev Pereira, R. Pugliesi
Fig. 1. Typical ceramic vessels studied in the present work.
Fig. 2. X-ray diffractogram of the studied ceramic.
Two vessels were prepared for tomography. They were kept for
seven days at atmospheric pressure, at two air relative humidity (RH)
conditions, 40%, and 90%. After this period, the Paraloid® B-72 solution
was applied in its outer walls using a brush [7]. They were kept in the
laboratory (𝑇 = 24 ◦ C, 𝑅𝐻 = 64%) for 3 h to dry the acetone, prior the
irradiation takes place.
2.1. Ceramic characterization
2.1.1. Mineralogical composition
The composition was determined by X-ray diffraction (XRD) and for
such purpose, one of the vessels was crushed (∼30 μm grain size), in a
porcelain mortar, and a small fraction of the powder was analyzed in
a Rigaku (Multiflex) X-ray diffractometer. The result obtained in Fig. 2
shows that the ceramic mainly consists of Quartz (SiO2 ) and Orthoclase
(KAlSi3 O8 ) [7,10].
2.1.2. Ceramic surface
The surface was studied by Scanning Electronic Microscopy (SEM).
Two fragments of two vessels, the first without and the second with
Paraloid® B-72, were imaged in the electronic microscope FEI (mod.
QUANTA 400 FEG). The Fig. 3(a) is the image of the first fragment
showing some pores and cracks with irregular contours. The Fig. 3(b)
shows the second fragment in which the Paraloid® B-72 is visible as
a homogeneous film covering the ceramic surface, sealing some pores
and cracks, and partially sealing other ones, keeping its permeability to
humidity [5–7].
2.1.3. Radioactivity of the ceramic
This study was carried out in two steps:
– step one: according to [11,12], the raw-materials employed to man-
ufacture ceramics, usually contain naturally occurring radionuclides,
mainly U-238, Th-232 and K-40. In order to verify such occurrences
in the Macuxi ceramic, 17 g of a powdered non-irradiated vessel, was
analyzed in a high-resolution gamma spectroscope. The same radionu-
clides in concentrations very close to that ones, found in ordinary
contemporary red ceramics were observed.
119
Nuclear Inst. and Methods in Physics Research, A 889 (2018) 118–121
Fig. 3. Photographs of the ceramic fragments without (a) and with (b) Paraloid®
B-72.
Fig. 4. Sketch of the equipment for neutron tomography.
– step two: to perform the tomography in the present equipment, the
samples (ceramic vessels) are irradiated in an intense neutron beam (see
3.1), and radioactive activation of some of their chemical elements is ex-
pected. In order to determine the period, after irradiation, for which the
induced radioactivity becomes negligible, 0.7 g of the same powdered
vessel, was inserted in an aluminum holder and irradiated at the sample
position for 400 s, the required time to obtain a tomography [13]. The
powder was analyzed in a high sensitivity Ludlum (mod. 2929 dual
scaler) counting system, which is able to detect alpha, beta and gamma
radiations. The obtained result showed that approximately 30 h after the
end of the irradiation, the level of counting of the irradiated ceramics
became insignificant, as it reached the background level of counting
system [14].
3. Penetration and distribution of paraloid® B-72
3.1. Neutron tomography
The equipment for tomography is installed at the Beam–Hole (BH)
#14, of the 5MW IEA-R1 Nuclear Research Reactor, of IPEN-CNEN/SP
(Brazilian Institute for Nuclear Technology). Fig. 4 shows the sketch of
its main components: the rotating table where the vessels are irradiated;
the scintillator screen (Li6 F/ZNS-NE426) where the image is formed;
the plane mirror to reflect the image of the scintillator to the video
camera; the CCD video camera (ANDOR ikon-M) for imaging capture;
the radiation shielding. Furthermore two softwares, installed in a DELL
(precision 5500) work station are used: Octopus V8.0 that from the
captured images, generates tomographs (or slices) in the three viewing
planes (𝑋𝑌 , 𝑋𝑍, 𝑌 𝑍), and VG Studio Max V2.2, which from the
tomographs provides 3D images of the inspected vessel [15,16]. The
neutron flux at irradiation position (measured by the Au-foil method)
is 8 ×106 n.s−1 ⋅ cm−2 , the maximal beam diameter is 16 cm, the best
achievable spatial resolution is 205 ± 25 μm, the field of view is 18
× 18 cm2 , the L/D ratio is 104 ± 4, and the time spent per tomography
is about 400 s. The Ref. [13] describe in detail, the equipment, and the
procedures for tomography.
As already mentioned, two vessels, in which the consolidant was
applied, at 40% RH and 90% RH, were tomographed. The penetration
of Paraloid® B-72 was evaluated from the brightness level intensity
M.A. Stanojev Pereira, R. Pugliesi Nuclear Inst. and Methods in Physics Research, A 889 (2018) 118–121

Fig. 5. Tomography slice — 128 for the vessel at 90% RH.

Table 1 Fig. 6. Typical gray level intensity distribution for the vessel at 90% RH.
Penetration of Paraloid® B-72 in the ceramic vessels.
Penetration of Paraloid® B-72 90% RH 40% RH
Depth at the highest concentration — (B–A) 1.1 ± 0.3 mm* 0.9 ± 0.3 mm**
Deepest penetration — (C–A) 3.4 ± 0.3 mm 3.9 ± 0.3 mm
* Gray Level 𝐷 = 36 ± 3.
**
Gray Level 𝐷 = 33 ± 3.

distributions, obtained in a selected slice of the tomography, as a

function of the scanning coordinate ‘‘𝑥’’ along to the perpendicular
direction to the vessel wall. The brightness level intensity was quantified
in an 8 bit Gray Level (GL) scale ranging from 0 (darkest level) to 255
(brightest level), in such way that the greater is the (GL) value, the
greater is the Paraloid® B-72 concentration [13].
Since the variations in the grain size, pore size, and the non-
uniformity of the drying process affect the penetration of Paraloid® B-72
in the ceramic material, several distributions in different positions of the
same slice, were determined.
For the vessel at 90% RH, twenty distributions in the slice 128,
corresponding to the viewing plane 𝑋𝑌 , were obtained. Fig. 5 is a 3D
image of this vessel showing the slice 128 at 3.5 cm from its surface, the
Paraloid® B-72 is the semi-circular red ring at left, the ceramic body
Fig. 7. Gray level distributions (black and blue) and the resulting net distribu-
is the light brown, and the arrows indicate the directions in which the
tion (red) for the vessel at 90% RH.
scans were obtained. A typically distribution is illustrated in Fig. 6 and
consists of two distinct parts, the first to ‘‘Ceramic wall + Paraloid’’
and the second to ‘‘Ceramic wall’’. The twenty distributions were
superimposed and the values of GL for each scanning coordinate were 3.2. 3D-neutron images
averaged. In this new distribution these two parts, were subtracted and
the resulting net distribution of Paraloid® B-72 penetration in the vessel 3D images have been applied since they provide a macro-
visualization of the distribution of Paraloid® B-72 in the ceramic body.
is shown in Fig. 7, which can be described as follows: the coordinate
Firstly and for reference, Fig. 10(a) shows the 3D image of the entire
(A) represents the beginning of the vessel wall, where the Paraloid®
vessel, as it is seen by the neutron beam. Fig. 10(b) and (c) show the
B-72 was applied; the concentration of the consolidant continuously
consolidant distribution along the viewing plane 𝑋𝑌 in two regions,
increases reaching its maximal at (B), for the Gray Level (D); as the
located at 13 mm and 18 mm from the vessel surface, corresponding
depth increases, the concentration continuously decreases, reaching the
to the slices 48 and 64 respectively. From these images, and because
deepest penetration at (C). Thus (B–A) is the depth in the vessel wall in
of the high sensitivity of neutrons to hydrogen (which is present in
which the Paraloid® B-72 reaches the maximal concentration and (C–A) the Paraloid® B-72), it was possible to visualize the consolidant in
is the deepest penetration. The obtained results are shown in Table 1. For the ceramic vessel, identify non-homogeneous impregnation regions,
the vessel at 40% RH, the same procedures were applied. The obtained failure or regions in which Paraloid® B-72 was not absorbed, among
distributions are illustrated in Figs. 8 and 9, and the respective results others [17].
in Table 1.
The uncertainties in A, B, C and D are the mean standard deviations 4. Conclusions
of their respective values corresponding to each one of the twenty
individual distributions, and the uncertainties in (B–A) and (C–A) were The results of Table 1 show that the penetration of Paraloid® B-72 in
evaluated by the standard propagation rules. Macuxi ceramic is not affected by the humidity conditions of the vessels.

120
M.A. Stanojev Pereira, R. Pugliesi Nuclear Inst. and Methods in Physics Research, A 889 (2018) 118–121

The thicknesses of the vessel walls vary between 5 mm to 9 mm and

since the maximal penetration of Paraloid® B-72 is about 3.4 ± 0.2 mm
for the vessel kept at 90% RH and of 3.9 ± 0.2 mm for the vessel at
40% RH, no part of the vessels is entirely protected.
Fig. 10 demonstrated the importance of 3D images for the present
study since they allow a macro-visualization of the distribution of
Paraloid® B-72 in the ceramic body, in any region of the three viewing
planes ‘‘𝑋𝑌 ’’, ‘‘𝑋𝑍’’, ‘‘𝑌 𝑍’’, or eventually in the entire vessel at once.
Another very important aspect verified was the radioactivity of
the ceramic. The results showed although the raw-materials contain
naturally occurring radionuclides, the neutron-induced radioactivity
becomes negligible approximately 30 h after the end of the tomography
procedure.
Considering that the studied vessels were manufactured by an indige-
nous artisan according to the same procedures used by their ancestors,
the obtained results can serve as a guide to extrapolate the behavior of
Paraloid® B-72 to other archeological ceramic objects from this same
region.

Acknowledgments
Fig. 8. Typical gray level intensity distribution for the vessel at 40% RH.
The authors are indebted to the Macuxi artisan Ms. Lidia Raposo, to
International Atomic Energy Agency — IAEA and to National Council
of Technological and Scientific Development — CNPq (Brazil) for the
postdoctoral fellowship partial financial support to this project through
the grant CRP code F11018 and the Fellowship 114862/2015-0 respec-
tively.

References

[1] Nuclear techniques for cultural heritage research, IAEA Radiation Technology Series
No. 2.
[2] O.A. Derby, The artificial mounds of the Island of Marajo, Brazil, Amer. Nat. 13 (4)
(1879) 224–229.
[3] F.A. Silva, Loiça de barro do agreste: um estudo etnoarqueológico de cerâmica
histórica pernambucana (M.Sc. thesis), University of São Paulo, Brazil, 2012.
[4] C. Constâncio, L. Franco, A. Russo, C. Anjinho, J. Pires, M.F. Vaz, A.P. Carvalho,
Studies on polymeric conservation treatments of ceramic tiles with Paraloid B-72
and two alkoxysilanes, J. Appl. Polym. 116 (2010) 2833–2839.
[5] G. Fleischer, J. Nimmrichter, A. Rohatsch, The relevance of scientific investigation
for the preservation of monuments and historic buildings made of stone e a state
of the art report, Geophys. Res. Abstr. 7 (2005) SRef-ID: 1607e7962/gra/EGU05-A-
04479.
[6] E. Carreti, L. Dei, Physicochemical characterization of acrylic polymeric coating
Fig. 9. Gray level distributions (black and blue) and the resulting net distribu- porous materials of artistic interest, Prog. Org. Coat. 49 (2004) 282–289.
tion (red) for the vessel at 40% RH. [7] M.I. Prudêncio, M.A. Pereira Stanojev, J.G. Marques, M.I. Dias, L. Esteves, C.I.
Burbidge, M.J. Trindade, M.B. Albuquerque, Neutron tomography for the assessment
of consolidant impregnation efficiency in portuguese glazed tiles (16th and 18th
centuries), J. Archaeol. Sci. 39 (2012) 964–969.
[8] T.P. Santos, M.F. Vaz, M.L. Pinto, A.P. Carvalho, Porosity characterization of old
Portuguese ceramic tiles, Constr. Build. Mater. 28 (2012) 104–110.
[9] https://pib.socioambiental.org/pt/povo/macuxi/734. (Accessed August 2017).
[10] M.F. Vaz, J. Pires, A.P. Carvalho, Effect of impregnation treatment with Paraloid
B-72 on the properties of old Portuguese ceramic tiles, J. Cult. Heritage 9 (2008)
269–276.
[11] http://nvlpubs.nist.gov/nistpubs/Legacy/TN/nbstechnicalnote1118.pdf. (Accessed
July 2016).
[12] http://nucleardata.nuclear.lu.se/toi/perchart.htm. (Accessed February 2017).
[13] R. Schoueri, C. Domienikan, F. Toledo, M.L.G. Andrade, M.A. Pereira Stanojev, R.
Pugliesi, The new facility for neutron tomography of IPEN-CNEN/SP and its potential
to investigate hydrogenous substances, Appl. Radiat. Isot. 84 (02/2014) (2014) 22–
26.
[14] Nuclear Energy National Commission, Basic Guidelines for Radiation Protection,
CNEN, Rio de Janeiro, 2014 (CNEN NN 3.01).
[15] https://www.octopus.be/nl. (Accessed September 2017).
[16] https://www.volumegraphics.com/. (Accessed September 2017).
[17] J. Rant, Z. Milic, J. Istenic, T. Knific, I. Lengar, A. Rant, Neutron radiography
examination of objects belonging to the cultural heritage, Appl. Radiat. Isot. 64 (1)
(2006) 7–12.
Fig. 10. 3D neutron images: (a) entire vessel; (b) and (c) distribution of the
consolidant in two distinct regions of the ceramic vessel.

121